Streaming is conventionally performed in a network on media data for video, TV, sound or multimedia content, and enables the playback of such data without the need to download the entire resource file in advance of playback. This means that such data has to be sent at a pace corresponding, at least on average, to the pace at which it is to be presented. In many streaming applications, users, e.g. video viewers, have display devices, e.g. television sets or personal computers, that are connectable, e.g. via an IP-network, to a server in which the media data is stored. Each user can request individually a video or a TV program stored in the server to be displayed, so that the video is displayed to the user immediately upon streaming from the server, i.e. without storage at the display device of the entire video or TV program. The stream is normally compressed or coded, for example by MPEG-2, and the display device is adapted to decode it.
It is known, for example from the patent documents U.S. Pat. No. 5,898,456, WO2004/049717A3 or WO00/60861, to distribute media data between servers in a network based on the popularity of contents in the media data, such that content with a high popularity is stored in servers that are close to the users, while content with a low popularity is stored in remote servers. Thereby, a content management unit is provided to gather information corresponding to the popularity of media contents for streaming in the network, and to manage the distribution and re-distribution of such contents in servers in the network. However, this requires a large amount of data traffic with media popularity and management information to be exchanged in the network, and such data traffic will require a large amount of the bandwidth available in the network. Also, the re-distribution of media data contents between the servers requires a relatively large amount of the bandwidth available in the network.
There is also a need to meet demands from users for a large variety of media data contents, and at the same time to keep memory capacity requirements of streaming servers within reasonable and cost effective boundaries.
There is further a need to decrease the amount of transfers of media data in a network in order to increase the efficiency in the use of the network bandwidth capacity.